Crystal nucleation in ionic and covalent systems

Most liquids can be cooled below their equilibrium freezing point before crystals start to form in the bulk. Crystal growth can only proceed when, due to some rare fluctuation, one or more micro-crystallites exceed a critical size (the ''critical nucleus''). This process is called homogeneous nucleation.

At moderate super-cooling, nucleation does not happen spontaneously, and it is too slow to be studied with ''brute-force'' Molecular Dynamics or Monte Carlo simulations. Therefore, novel ''rare event techniques'' are needed to bridge the time-scale gap between microscopic dynamics and the time it takes for a (rare) nucleation event to take place.

In my talk, I will focus on numerical simulation results of homogeneous nucleation of two systems of our interest: diamond nucleation from molten carbon and crystal nucleation of NaCl from its melt.

While in the diamond nucleation, we demonstrate that the structure of a network-forming liquid is crucial for the rate of crystallization, in the NaCl we observe that the critical nucleus formed during the nucleation process, not only has the crystal structure of bulk NaCl, but it is also clearly facetted, showing a cubical shape. We also discuss the discrepancy between the interfacial free-energy obtained from simulations of the crystal nucleation barrier with those obtained from numerical studies of the solid-liquid contact-angle near the triple point in the sodium chloride case.